Practical Social Analytics

© 2008 Critical Mass, Inc. All Rights Reserved

Practical Social Analytics@ryancoleman This seems relevant to my interests.

June 12, 2009

2© 2008 Critical Mass, Inc. All Rights Reserved

Three Core Takeaways

• We’ve seen this before• Understand the hype• Move beyond it to where it’s practical

• The Medium is the Message• It’s people

• Some of it is usefully measurable • Revise your expectations• You can’t measure it all (affordably)• Start with your goals• Worry about the tools and ‘how’ later


We’ve Seen This Before


This isn’t new…

“Since opinion leaders are specialized by groups and type of product, a marketing research should not attempt to find opinion leaders “in general”. Rather he should first look for the various strata among users of the particular product, and then determine the opinion leaders in each stratum.”







Social media has amplified the effect of WOM in a major way


For Better and For Worse


Hype Cycle



Takeaway

We’ve seen this before.

It’s alright to be excited.

This is no excuse to forget everything you already know about goals and measurability

Think back to what you already know about measurement


The Medium and The Message


The Media isn’t the Medium

Regular presentations stop there: they talk about the medium.

About social media.

We’re so focused on the Mediums...

To the original point – about Word of Mouth (WOM)

What’s the Medium?


It’s people


And How People Organize Each Other?

Hierarchies exist where there is ambiguity in transactions, resulting in the need for authority. Authority manifests itself in the form of hierarchy.

Networks are constructive and destructive because they transcend hierarchy. They operate on the basis of trust.

This is a human design pattern.

-Karen Stevenson


Hierarchy


Networks – Who would you target first?

Degrees of Trust


Networks and Hierarchies

The key variable Social Measurement should be

Trust


People are The Medium and Trust is the Currency

Trust is the independent variable that has a massive influence on:

• Credibility• Likelihood of a message to be repeated / linked• Affinity

And can even impact satisfaction

How likely are you to be satisfied with a restaurant that recently gave your friend food poisoning?



Takeaway

The Medium is the Message.

The Medium is People.

The Message is People.

When confronted with the awesome complexity of social analytics, remember:

It’s about People.

It should be about “Trust”…but… it’s not practical.


PRACTICAL

Social

Analytics


This isn’t 2Unlimited


There are most certainly are limits


Costs


Privacy






Social Measurement Strategy

A comprehensive social measurement strategy includes:

1.Volumetric tracking of conversations · What are people talking about?

2.Affinity / Sentiment measurement · How are people saying it?

3.Social Graph Measurement • Do the people saying it matter? Who do we need to

reach out to, to make it happen?• Do they trust each other?

4.Monetization • How, roughly, is it impacting our bottom line, brand


The whole product

The four components, combined, is the whole ‘social analytics’ product.

No single vendor has ‘the whole product’. While many vendors accurately code ‘sentiment’ there is denial of the importance of social graphing.

The best we can do, in the interim, is to insist on the interoperability of the toolsets.


1. Volumetric

Volumetric tracking of conversations (What are people talking about?)

Typical Metrics:

1. Number of Blogger Mentions in x Period (Source: Organic Search Results Google: filtering for Blogger)

2. Number of Tweet Mentions in x Period (Source: Free Search Option)

3. Number of Unique Author Mentions in x Period (Source: Manual hand count. Ie. You may have gotten 15 tweets, but were 10 of them from the same fanatical person?)

4. Number of Diggs in x Period (Source: Digg, certain social bookmarking aggregators)

…and so on


Consider Twitter, For Instance


Tweet Stats…


More Tweetstats


Trends


You can use Google to do a count of mentions…


And you can sort by TIME too!


They can even be derived from YouTube!


There’s all the regular data…



Affinity / Sentiment Measurement

Affinity / Sentiment measurement (How are people saying it?)

Typical Metrics:

1. Count of Positive Comments in x Period (Vendor provided or Intern Coded)

2. Count of Negative Comments in x Period (Vendor provided or Intern Coded)

3. Count of Constructive Comments in x Period (Vendor provided or Intern Coded)

4. Count of Recommendation Comments in x Period (Vendor provided or Intern Coded)

5. The Emotional Connection of a Comment to the Topic (Vendor provided)

…and so on


Guess the Sentiment


Social Graph Measurement

Social Graph Measurement (Do the people saying it matter? Who do we need to reach out to, to make it happen?)

Typical Metrics:

1. Social Eigenvector (Source: Vendor Provided)

2. Hubs (Source: Vendor Provided)

3. Influencers (Source: Vendor Provided)

4. Gatekeepers (Source: Vendor Provided)

5. The Psychographic Distance between the Commentor and the Commentee (Source: Vendor Provided)

6. Homophily / Homogeneity / Heterogeneity / Weak Ties (Source: Vendor Provided)

7. The Degree of Trust between You and Your Network (Source: Vendor Provided)

…and so on.


The Strength of Weak Ties

George

Penelope

Sky

Ocean Gertrude

Creek


And…there’s homebrew…Week 1


…Week 2


…Week 3


Monetization Measurement

Monetization (How, roughly, is it impacting our bottom line, brand)

Typical Metrics:

1. Revenue directly attributable to a social media campaign (Source: Web Analytics tagging (Google Analytics, for instance)

2. Revenue indirectly attributable to a social media campaign (Source: Survey, Correlation Analysis)


And Most Monetization Can Be Tracked…

Google Analytics*, if set up properly, can be instrumental in measuring directly attributable ROI from Social Campaigns

*And it’s “free”


That’s Nice


That’s nice, so what?

All of this ‘measurement’ is only useful if it actually aligns with your organizations goals


Most people start with the Tool

There are hundreds of tools out there.

Just because something can be measured doesn’t mean that it should.

Start with the goal. Work frontwards.

You’ll be less dissatisfied that way.


Grappling with Complexity

ASK:

1. What is the goal?

2. Which metrics are indicative or predictive of achieving the goal?

3. How can you probably measure it?

4. Can you actually measure it?

5. If not, is there a cheaper way of measuring it?

6. If so, measure it!


Three Core Takeaways

• We’ve seen this before• Understand the hype• Move beyond it to where it’s practical

• The Medium is the Message• It’s people

• Some of it is usefully measurable• Revise your expectations• You can’t measure it all (affordably)• Start with your goals• Worry about the tools and ‘how’ later


Contact

Questions?

Christopher Berry [email protected]

Twitter: cjpberry

#WA Blog: http://christopher-berry.blogspot.com/

Social Tech Blog: http://christopherberry.ca/

Practical Social Analytics

Technology